From: not_valid@nospicedham.comcast.net   
      
   "Bernhard Schornak"  wrote in message news:pj1ole$sl1$1@dont-email.me...   
      
   > fizz buzz wrote:   
      
   > Plain copy operations are faster than complex calculations with   
   > a lot of comparisons and not continuously repeating conditional   
   > jumps. One conditional branch with random branch target 'costs'   
   > the same time than multiple cacheline writes, shuffling 64 byte   
   > blocks simultaneously.   
      
   That's why you do 8 in parallel in ymm registers with branchless   
   code.   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)